CMOS image sensor with adjustable sensitivity is wildly adopted to control pixel conversion gain (CG) with respect to light intensity. High light intensity requires low conversion gain to avoid saturation. Low light intensity requires high conversion gain so signal can be read out with shorter charge transfer time. However, conventional CMOS image sensor with adjustable sensitivity consumes extra hardware for adjusting capacitance of the charge storage node. The extra hardware causes the fill factor to be reduced and a dilemma arises.